mirror of
https://github.com/mudler/LocalAI.git
synced 2025-05-20 02:24:59 +00:00
chore(parler-tts): drop backend (#4672)
We support at this point more extensive backends that are SOTA and support also voice cloning, and many other features. This backend is superseded and also poses significant maintenance burden as there is an open issue https://github.com/mudler/LocalAI/issues/3941 which is still open as it deps are pinning old versions of grpc. Closes https://github.com/mudler/LocalAI/issues/3941 Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
This commit is contained in:
Ettore Di Giacinto
GitHub
parent
89429a439b
commit
318225f631
16 changed files with 4 additions and 343 deletions
|
|
@ -1,44 +0,0 @@
|
|||
export CONDA_ENV_PATH = "parler.yml"
|
||||
SKIP_CONDA?=0
|
||||
ifeq ($(BUILD_TYPE), cublas)
|
||||
export CONDA_ENV_PATH = "parler-nvidia.yml"
|
||||
endif
|
||||
|
||||
# Intel GPU are supposed to have dependencies installed in the main python
|
||||
# environment, so we skip conda installation for SYCL builds.
|
||||
# https://github.com/intel/intel-extension-for-pytorch/issues/538
|
||||
ifneq (,$(findstring sycl,$(BUILD_TYPE)))
|
||||
export SKIP_CONDA=1
|
||||
endif
|
||||
|
||||
.PHONY: parler-tts
|
||||
parler-tts:
|
||||
@echo "Installing $(CONDA_ENV_PATH)..."
|
||||
bash install.sh $(CONDA_ENV_PATH)
|
||||
$(MAKE) protogen
|
||||
|
||||
.PHONY: run
|
||||
run: protogen
|
||||
@echo "Running transformers..."
|
||||
bash run.sh
|
||||
@echo "transformers run."
|
||||
|
||||
.PHONY: test
|
||||
test: protogen
|
||||
@echo "Testing transformers..."
|
||||
bash test.sh
|
||||
@echo "transformers tested."
|
||||
|
||||
.PHONY: protogen
|
||||
protogen: backend_pb2_grpc.py backend_pb2.py
|
||||
|
||||
.PHONY: protogen-clean
|
||||
protogen-clean:
|
||||
$(RM) backend_pb2_grpc.py backend_pb2.py
|
||||
|
||||
backend_pb2_grpc.py backend_pb2.py:
|
||||
bash protogen.sh
|
||||
|
||||
.PHONY: clean
|
||||
clean: protogen-clean
|
||||
$(RM) -r venv __pycache__
|
||||
|
|
@ -1,125 +0,0 @@
|
|||
#!/usr/bin/env python3
|
||||
"""
|
||||
Extra gRPC server for MusicgenForConditionalGeneration models.
|
||||
"""
|
||||
from concurrent import futures
|
||||
|
||||
import argparse
|
||||
import signal
|
||||
import sys
|
||||
import os
|
||||
|
||||
import time
|
||||
import backend_pb2
|
||||
import backend_pb2_grpc
|
||||
|
||||
import grpc
|
||||
|
||||
from scipy.io.wavfile import write as write_wav
|
||||
|
||||
from parler_tts import ParlerTTSForConditionalGeneration
|
||||
from transformers import AutoTokenizer
|
||||
import soundfile as sf
|
||||
import torch
|
||||
|
||||
_ONE_DAY_IN_SECONDS = 60 * 60 * 24
|
||||
|
||||
# If MAX_WORKERS are specified in the environment use it, otherwise default to 1
|
||||
MAX_WORKERS = int(os.environ.get('PYTHON_GRPC_MAX_WORKERS', '1'))
|
||||
|
||||
# Implement the BackendServicer class with the service methods
|
||||
class BackendServicer(backend_pb2_grpc.BackendServicer):
|
||||
"""
|
||||
A gRPC servicer for the backend service.
|
||||
|
||||
This class implements the gRPC methods for the backend service, including Health, LoadModel, and Embedding.
|
||||
"""
|
||||
def Health(self, request, context):
|
||||
"""
|
||||
A gRPC method that returns the health status of the backend service.
|
||||
|
||||
Args:
|
||||
request: A HealthRequest object that contains the request parameters.
|
||||
context: A grpc.ServicerContext object that provides information about the RPC.
|
||||
|
||||
Returns:
|
||||
A Reply object that contains the health status of the backend service.
|
||||
"""
|
||||
return backend_pb2.Reply(message=bytes("OK", 'utf-8'))
|
||||
|
||||
def LoadModel(self, request, context):
|
||||
"""
|
||||
A gRPC method that loads a model into memory.
|
||||
|
||||
Args:
|
||||
request: A LoadModelRequest object that contains the request parameters.
|
||||
context: A grpc.ServicerContext object that provides information about the RPC.
|
||||
|
||||
Returns:
|
||||
A Result object that contains the result of the LoadModel operation.
|
||||
"""
|
||||
model_name = request.Model
|
||||
device = "cuda:0" if torch.cuda.is_available() else "cpu"
|
||||
try:
|
||||
self.model = ParlerTTSForConditionalGeneration.from_pretrained(model_name).to(device)
|
||||
self.tokenizer = AutoTokenizer.from_pretrained(model_name)
|
||||
except Exception as err:
|
||||
return backend_pb2.Result(success=False, message=f"Unexpected {err=}, {type(err)=}")
|
||||
|
||||
return backend_pb2.Result(message="Model loaded successfully", success=True)
|
||||
|
||||
def TTS(self, request, context):
|
||||
model_name = request.model
|
||||
voice = request.voice
|
||||
if voice == "":
|
||||
voice = "A female speaker with a slightly low-pitched voice delivers her words quite expressively, in a very confined sounding environment with clear audio quality. She speaks very fast."
|
||||
if model_name == "":
|
||||
return backend_pb2.Result(success=False, message="request.model is required")
|
||||
try:
|
||||
device = "cuda:0" if torch.cuda.is_available() else "cpu"
|
||||
input_ids = self.tokenizer(voice, return_tensors="pt").input_ids.to(device)
|
||||
prompt_input_ids = self.tokenizer(request.text, return_tensors="pt").input_ids.to(device)
|
||||
|
||||
generation = self.model.generate(input_ids=input_ids, prompt_input_ids=prompt_input_ids)
|
||||
audio_arr = generation.cpu().numpy().squeeze()
|
||||
print("[parler-tts] TTS generated!", file=sys.stderr)
|
||||
sf.write(request.dst, audio_arr, self.model.config.sampling_rate)
|
||||
print("[parler-tts] TTS saved to", request.dst, file=sys.stderr)
|
||||
print("[parler-tts] TTS for", file=sys.stderr)
|
||||
print(request, file=sys.stderr)
|
||||
except Exception as err:
|
||||
return backend_pb2.Result(success=False, message=f"Unexpected {err=}, {type(err)=}")
|
||||
return backend_pb2.Result(success=True)
|
||||
|
||||
|
||||
def serve(address):
|
||||
server = grpc.server(futures.ThreadPoolExecutor(max_workers=MAX_WORKERS))
|
||||
backend_pb2_grpc.add_BackendServicer_to_server(BackendServicer(), server)
|
||||
server.add_insecure_port(address)
|
||||
server.start()
|
||||
print("[parler-tts] Server started. Listening on: " + address, file=sys.stderr)
|
||||
|
||||
# Define the signal handler function
|
||||
def signal_handler(sig, frame):
|
||||
print("[parler-tts] Received termination signal. Shutting down...")
|
||||
server.stop(0)
|
||||
sys.exit(0)
|
||||
|
||||
# Set the signal handlers for SIGINT and SIGTERM
|
||||
signal.signal(signal.SIGINT, signal_handler)
|
||||
signal.signal(signal.SIGTERM, signal_handler)
|
||||
|
||||
try:
|
||||
while True:
|
||||
time.sleep(_ONE_DAY_IN_SECONDS)
|
||||
except KeyboardInterrupt:
|
||||
server.stop(0)
|
||||
|
||||
if __name__ == "__main__":
|
||||
parser = argparse.ArgumentParser(description="Run the gRPC server.")
|
||||
parser.add_argument(
|
||||
"--addr", default="localhost:50051", help="The address to bind the server to."
|
||||
)
|
||||
args = parser.parse_args()
|
||||
print(f"[parler-tts] startup: {args}", file=sys.stderr)
|
||||
serve(args.addr)
|
||||
|
|
@ -1,28 +0,0 @@
|
|||
#!/bin/bash
|
||||
set -e
|
||||
|
||||
source $(dirname $0)/../common/libbackend.sh
|
||||
|
||||
# This is here because the Intel pip index is broken and returns 200 status codes for every package name, it just doesn't return any package links.
|
||||
# This makes uv think that the package exists in the Intel pip index, and by default it stops looking at other pip indexes once it finds a match.
|
||||
# We need uv to continue falling through to the pypi default index to find optimum[openvino] in the pypi index
|
||||
# the --upgrade actually allows us to *downgrade* torch to the version provided in the Intel pip index
|
||||
if [ "x${BUILD_PROFILE}" == "xintel" ]; then
|
||||
EXTRA_PIP_INSTALL_FLAGS+=" --upgrade --index-strategy=unsafe-first-match"
|
||||
fi
|
||||
|
||||
|
||||
installRequirements
|
||||
|
||||
|
||||
# https://github.com/descriptinc/audiotools/issues/101
|
||||
# incompatible protobuf versions.
|
||||
PYDIR=python3.10
|
||||
pyenv="${MY_DIR}/venv/lib/${PYDIR}/site-packages/google/protobuf/internal/"
|
||||
|
||||
if [ ! -d ${pyenv} ]; then
|
||||
echo "(parler-tts/install.sh): Error: ${pyenv} does not exist"
|
||||
exit 1
|
||||
fi
|
||||
|
||||
curl -L https://raw.githubusercontent.com/protocolbuffers/protobuf/main/python/google/protobuf/internal/builder.py -o ${pyenv}/builder.py
|
||||
|
|
@ -1,6 +0,0 @@
|
|||
#!/bin/bash
|
||||
set -e
|
||||
|
||||
source $(dirname $0)/../common/libbackend.sh
|
||||
|
||||
python3 -m grpc_tools.protoc -I../.. --python_out=. --grpc_python_out=. backend.proto
|
||||
|
|
@ -1,4 +0,0 @@
|
|||
git+https://github.com/huggingface/parler-tts.git@8e465f1b5fcd223478e07175cb40494d19ffbe17
|
||||
llvmlite==0.43.0
|
||||
numba==0.60.0
|
||||
grpcio-tools==1.42.0
|
||||
|
|
@ -1,3 +0,0 @@
|
|||
transformers
|
||||
accelerate
|
||||
torch==2.4.1
|
||||
|
|
@ -1,5 +0,0 @@
|
|||
--extra-index-url https://download.pytorch.org/whl/cu118
|
||||
torch==2.4.1+cu118
|
||||
torchaudio==2.4.1+cu118
|
||||
transformers
|
||||
accelerate
|
||||
|
|
@ -1,4 +0,0 @@
|
|||
torch==2.4.1
|
||||
torchaudio==2.4.1
|
||||
transformers
|
||||
accelerate
|
||||
|
|
@ -1,5 +0,0 @@
|
|||
--extra-index-url https://download.pytorch.org/whl/rocm6.0
|
||||
torch==2.3.0+rocm6.0
|
||||
torchaudio==2.3.0+rocm6.0
|
||||
transformers
|
||||
accelerate
|
||||
|
|
@ -1,8 +0,0 @@
|
|||
--extra-index-url https://pytorch-extension.intel.com/release-whl/stable/xpu/us/
|
||||
intel-extension-for-pytorch==2.3.110+xpu
|
||||
torch==2.3.1+cxx11.abi
|
||||
torchaudio==2.3.1+cxx11.abi
|
||||
oneccl_bind_pt==2.3.100+xpu
|
||||
optimum[openvino]
|
||||
transformers
|
||||
accelerate
|
||||
|
|
@ -1,4 +0,0 @@
|
|||
grpcio==1.69.0
|
||||
certifi
|
||||
llvmlite==0.43.0
|
||||
setuptools
|
||||
|
|
@ -1,4 +0,0 @@
|
|||
#!/bin/bash
|
||||
source $(dirname $0)/../common/libbackend.sh
|
||||
|
||||
startBackend $@
|
||||
|
|
@ -1,81 +0,0 @@
|
|||
"""
|
||||
A test script to test the gRPC service
|
||||
"""
|
||||
import unittest
|
||||
import subprocess
|
||||
import time
|
||||
import backend_pb2
|
||||
import backend_pb2_grpc
|
||||
|
||||
import grpc
|
||||
|
||||
|
||||
class TestBackendServicer(unittest.TestCase):
|
||||
"""
|
||||
TestBackendServicer is the class that tests the gRPC service
|
||||
"""
|
||||
def setUp(self):
|
||||
"""
|
||||
This method sets up the gRPC service by starting the server
|
||||
"""
|
||||
self.service = subprocess.Popen(["python3", "backend.py", "--addr", "localhost:50051"])
|
||||
time.sleep(10)
|
||||
|
||||
def tearDown(self) -> None:
|
||||
"""
|
||||
This method tears down the gRPC service by terminating the server
|
||||
"""
|
||||
self.service.terminate()
|
||||
self.service.wait()
|
||||
|
||||
def test_server_startup(self):
|
||||
"""
|
||||
This method tests if the server starts up successfully
|
||||
"""
|
||||
try:
|
||||
self.setUp()
|
||||
with grpc.insecure_channel("localhost:50051") as channel:
|
||||
stub = backend_pb2_grpc.BackendStub(channel)
|
||||
response = stub.Health(backend_pb2.HealthMessage())
|
||||
self.assertEqual(response.message, b'OK')
|
||||
except Exception as err:
|
||||
print(err)
|
||||
self.fail("Server failed to start")
|
||||
finally:
|
||||
self.tearDown()
|
||||
|
||||
def test_load_model(self):
|
||||
"""
|
||||
This method tests if the model is loaded successfully
|
||||
"""
|
||||
try:
|
||||
self.setUp()
|
||||
with grpc.insecure_channel("localhost:50051") as channel:
|
||||
stub = backend_pb2_grpc.BackendStub(channel)
|
||||
response = stub.LoadModel(backend_pb2.ModelOptions(Model="parler-tts/parler_tts_mini_v0.1"))
|
||||
self.assertTrue(response.success)
|
||||
self.assertEqual(response.message, "Model loaded successfully")
|
||||
except Exception as err:
|
||||
print(err)
|
||||
self.fail("LoadModel service failed")
|
||||
finally:
|
||||
self.tearDown()
|
||||
|
||||
def test_tts(self):
|
||||
"""
|
||||
This method tests if the embeddings are generated successfully
|
||||
"""
|
||||
try:
|
||||
self.setUp()
|
||||
with grpc.insecure_channel("localhost:50051") as channel:
|
||||
stub = backend_pb2_grpc.BackendStub(channel)
|
||||
response = stub.LoadModel(backend_pb2.ModelOptions(Model="parler-tts/parler_tts_mini_v0.1"))
|
||||
self.assertTrue(response.success)
|
||||
tts_request = backend_pb2.TTSRequest(text="Hey, how are you doing today?")
|
||||
tts_response = stub.TTS(tts_request)
|
||||
self.assertIsNotNone(tts_response)
|
||||
except Exception as err:
|
||||
print(err)
|
||||
self.fail("TTS service failed")
|
||||
finally:
|
||||
self.tearDown()
|
||||
|
|
@ -1,6 +0,0 @@
|
|||
#!/bin/bash
|
||||
set -e
|
||||
|
||||
source $(dirname $0)/../common/libbackend.sh
|
||||
|
||||
runUnittests
|
||||
Loading…
Add table
Add a link
Reference in a new issue